Next Article in Journal
Improving the Performance of Continuous-Variable Measurement-Device-Independent Quantum Key Distribution via a Noiseless Linear Amplifier
Previous Article in Journal
A Smooth Path between the Classical Realm and the Quantum Realm
Previous Article in Special Issue
Gradient Regularization as Approximate Variational Inference
Article

Perfect Density Models Cannot Guarantee Anomaly Detection

by 1,2,* and 2
1
Department of Statistics, University of Oxford, Oxford OX1 3LB, UK
2
Google Research, Montreal, QC H3B 2Y5, Canada
*
Author to whom correspondence should be addressed.
Academic Editors: Eric Nalisnick and Boris Ryabko
Entropy 2021, 23(12), 1690; https://doi.org/10.3390/e23121690
Received: 30 September 2021 / Revised: 3 November 2021 / Accepted: 13 December 2021 / Published: 16 December 2021
(This article belongs to the Special Issue Probabilistic Methods for Deep Learning)
Thanks to the tractability of their likelihood, several deep generative models show promise for seemingly straightforward but important applications like anomaly detection, uncertainty estimation, and active learning. However, the likelihood values empirically attributed to anomalies conflict with the expectations these proposed applications suggest. In this paper, we take a closer look at the behavior of distribution densities through the lens of reparametrization and show that these quantities carry less meaningful information than previously thought, beyond estimation issues or the curse of dimensionality. We conclude that the use of these likelihoods for anomaly detection relies on strong and implicit hypotheses, and highlight the necessity of explicitly formulating these assumptions for reliable anomaly detection. View Full-Text
Keywords: deep generative modeling; probabilistic modeling; anomaly detection deep generative modeling; probabilistic modeling; anomaly detection
Show Figures

Figure 1

MDPI and ACS Style

Le Lan, C.; Dinh, L. Perfect Density Models Cannot Guarantee Anomaly Detection. Entropy 2021, 23, 1690. https://doi.org/10.3390/e23121690

AMA Style

Le Lan C, Dinh L. Perfect Density Models Cannot Guarantee Anomaly Detection. Entropy. 2021; 23(12):1690. https://doi.org/10.3390/e23121690

Chicago/Turabian Style

Le Lan, Charline, and Laurent Dinh. 2021. "Perfect Density Models Cannot Guarantee Anomaly Detection" Entropy 23, no. 12: 1690. https://doi.org/10.3390/e23121690

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop